natural hazards 2015

Latest on the Alpine Fault

NATURAL HAZARDS 2015

RiskScape at Home & Abroad

Platform’s New Research Projects

NATURAL HAZARDS | 2015

Cover credits Front: Pilar Villamor (right, GNS Science) and PhD student Monica Giona Bucci (left, Lincoln University) examine liquefaction at Pines Beach, Christchurch following the earthquake on 14 February 2016. Back: Julian Thomson (GNS Science) controls the drone overhead at Pines Beach. Drone images allow scientists to quickly survey vast areas of land for signs of damage.

Citation Pinal, C and Brackley, H. (Eds.) 2016. Natural Hazards 2015. Lower Hutt, NZ: GNS Science. GNS Science Miscellaneous Series 88, 44 p. Our thanks to Eileen McSaveney (GNS Science) and Colin Barkus (NIWA) for their editorial assistance.

GNS Science Miscellaneous Series 88 ISSN 1177-2441 (Print) ISSN 1172-2886 (Online) ii

NHRP

CONTENTS FOREWORD.................................................................................................................................... 2 PLATFORM MANAGER’S PERSPECTIVE....................................................................................... 3 PAST LARGE EARTHQUAKES ON THE ALPINE FAULT ............................................................... 5 UNDERSTANDING MODERN LIQUEFACTION TO UNDERSTAND THE PAST............................ 10 Student Profile: Monica Giona Bucci....................................................................................................... 14

INTRODUCING THE PLATFORM’S NEW PROJECTS: PARTNER-LED....................................... 15 INTRODUCING THE PLATFORM’S NEW PROJECTS: CONTEST 2015....................................... 16 UPDATES FROM THE CANTERBURY EARTHQUAKE RESEARCH PROGRAMME.................... 18 SQuADS: Subjective-Quantitative Assessment Decision Support.......................................................... 18 Rethinking Probabilistic Seismic Hazard Assessment............................................................................ 19

NEW ZEALAND’S RISKSCAPE..................................................................................................... 20 Fragility Functions: Key to Tsunami Risk Assessments.........................................................................22 RiskScape in Chile ....................................................................................................................................24 Post-Event Flood Survey: Whanganui ....................................................................................................26 Developing RiskScape for Volcanic Hazards...........................................................................................28 Quantifying New Zealand’s Coastal Risk Exposure................................................................................30

THE EFFECT OF THE UNDERLYING SOIL ON EARTHQUAKE RESPONSE............................... 32 IMPROVING NEW ZEALAND’S RESILIENCE TO WIND STORMS............................................... 36 DESIGNING A BUILDING FOR WIND AND EARTHQUAKES ...................................................... 39 NATURAL HAZARDS IN 2015....................................................................................................... 40

1


FOREWORD There’s never really a quiet time in New Zealand when it comes to natural hazards and 2015 has been no exception. With several earthquakes − the M6.0 Wilberforce earthquake in January and the M8.3 Chile earthquake/tsunami-threat in September which resulted in tsunami warnings being issued for New Zealand coastlines; weather-related events − Tropical Cyclone Pam in March; significant flood events over May-June in Wellington, Dunedin, the West Coast, Taranaki, Manawatu and Whanganui, and dry conditions leading to drought in the parts of the South Island − this has been illustrated all too clearly. The response to and recovery from these events again demonstrates the importance of working together in times of emergency by the people involved in the science community, emergency services, local and central government, lifeline utilities, private and not-for-profit sectors. It also emphasises the on-going need to assess our natural hazards and the risks from those hazards both pre- and post-event. A key priority for the next two years is the development of a National Disaster Resilience Strategy. The current Strategy has guided effective Civil Defence Emergency Management (CDEM) for almost 14 years, resulting in solid emergency response arrangements, increased recognition of CDEM as a profession and improved integration of activities across stakeholders. However strong our response efforts are, New Zealand is still faced with an increased awareness of its hazards and the effects that these can have on our communities.

There are significant opportunities to strengthen New Zealand’s ability to minimise the consequences of disasters on our communities. The shocks, crises, and emergencies that New Zealand will inevitably face do not need to become ‘disasters’ that compromise our prosperity and living standards. International best practise suggests that for New Zealand to achieve its vision of resilience, a collective effort should shift the focus to ‘managing risk’ rather than ‘managing disasters’. In March 2015 New Zealand made a commitment to the international Sendai Framework for Disaster Risk Reduction. Within 15 years the Framework seeks to achieve: “The substantial reduction of disaster risk and losses in lives, livelihoods and health and in the economic, physical, social, cultural and environmental assets of persons, businesses, communities and countries.” The intent for New Zealand is that we examine our current work and consider where efforts could be better targeted to yield the greatest benefit across four priority areas: 1.

Understanding disaster risk;

2.

Strengthening governance to manage disaster risk;

3.

Investing in disaster risk reduction for resilience;

4.

Enhancing disaster preparedness for effective response, recovery, rehabilitation and reconstruction.

This is a unique opportunity for us to deepen relationships between the science community, local and central government, the private and not-for-profit sectors and, most importantly, with our communities to think ambitiously about how we can all contribute to building a resilient New Zealand. I hope you enjoy this edition of Natural Hazards 2015. The coming year promises to be no less busy for everyone involved in natural hazards research and disaster risk reduction, and I look forward to working with many of you throughout the year, including through the development of the National Disaster Resilience Strategy.

Sarah Stuart-Black

Director Ministry of Civil Defence & Emergency Management 2

NHRP

PLATFORM MANAGER’S PERSPECTIVE Members of the Platform Management Group (l-r): Peter Benfell, Opus Research; Sam Dean, NIWA; Gill Jolly, GNS Science; Hannah Brackley, Platform Manager; Peter Kemp, Massey University. Absent: Pierre Quenneville, University of Auckland; Rajesh Dhakal, University of Canterbury

Welcome to another edition of Natural Hazards. You may have noticed our new Platform branding for this issue, as well as on our website and other material. We’ve given ourselves a bit of a refresh and hope you like the new look!

In September 2015, the Platform signed a Memorandum of Understanding with the Australian Bushfire and Natural Hazards Cooperative Research Centre (BNHCRC), cementing a relationship that had developed over the previous year. Discussions are under way to share the research findings between Australia and New Zealand; this MoU will make that easier.

During 2015 we contracted six new partner-led research programmes that will run through to 2019 (see page 15). Further details of these can be found on our website, and we look forward to highlighting some of them in the next issue of Natural Hazards.

Like the Platform, the BNHCRC is a National Committee for the Integrated Research on Disaster Risk (IRDR). Both of us support and supplement IRDR’s research initiatives, and help to further develop links between national disaster risk reduction programmes within an international framework.

Research is well underway for the thirteen contestable projects that were also awarded in 2015 (see page 16) and one of these projects is featured in this issue (see page 32). Our next contestable funding round will be in early 2017 - keep an eye out for announcements. This year saw two new members join the Platform Management Group: Dr Sam Dean, Chief Scientist from NIWA and Professor Rajesh Dhakal from University of Canterbury. We formally welcome them both.

This year the Platform has continued its strong engagement with both central and local government agencies, supporting them in order to help New Zealand make progress towards the priorities of the Sendai Framework for Disaster Risk Reduction. The recent convergence of the Sendai Framework, Sustainable Development Goals and the Paris Climate Agreement has also produced an unprecedented opportunity to maximise the contribution of science-based disaster risk management to sustainable development.

The end of 2015 saw Kelvin Berryman step down from his role as Platform Director. His effective leadership and commitment to the Platform since its inception have been instrumental to the success of the Platform, and he left large shoes to fill. Although Kelvin is now in a new role at GNS Science, he remains strongly committed to natural hazard research working to best effect for New Zealand, the best team approach and the Platform model. Lastly, it is really encouraging to reflect on the contributions the Platform and other funding bodies have made towards developing future talent. Several of the articles in this issue feature research contributions by graduate students: from Lincoln University, Monica Giona Bucci (paleoliquefaction), and from University of Canterbury, James Williams (Chile tsunami), Daniel Blake (volcanic hazards), Rebecca Fitzgerald (volcanic hazards) and George Williams (volcanic hazards). Well done to all, and we look forward to seeing more from them in the future.

Hannah Brackley

Manager Natural Hazards Research Platform 3


ult a F ine p l A

John O’Groats wetland

4

ALPINE FAULT

PAST LARGE EARTHQUAKES ON THE ALPINE FAULT BY URSULA COCHRAN & KATE CLARK, GNS SCIENCE Contact: [email protected]

A future large earthquake on the Alpine Fault is inevitable. Now modern techniques are enabling the development of long earthquake records over the previous millennia. This ability to obtain the ages of past earthquakes means we can better forecast the timing of the next one and encourage appropriate preparations to ‘Get Ready, Get Thru’.

Aerial view of the Alpine Fault extending offshore from the John O’Groats wetland. Photo: Lloyd Homer.


A depiction of the Alpine Fault as it traverses the South Island. The area indicated by squares (top left) is shown in close-up (top right). An image of the John O’Groats River Valley (bottom, right) looking northeast along the Alpine Fault. Circles mark the core sites used in this research. Photo: GNS Science.

What we knew before long earthquake records:

We are here

Time (years) 6000

5000

4000

3000

2000

1000

BC/AD

1000

2000

Haast Earthquakes

Timelines showing ages of earthquakes (red dots) from studies at Haast (above) and Hokuri Creek and John O’Groats Wetland (next page).

What we know now: 6

We are here

Time (years) 6000

5000

4000

3000

2000

1000

BC/AD

1000

2000

ALPINE FAULT

Examples of the peat-silt couplets that represent earthquakes in cores at this site. Data: GNS Science.

The Significance of Long Earthquake Records

Coring a Wetland

Improved Long Earthquake Record

We went to John O’Groats River to look for data that could provide a more complete We compared the John O’Groats wetland record of activity over the last thousand record with that from Hokuri Creek, and years and improve our estimates of found that the John O’Groats site preserved earthquake frequency for the Southevidence of earthquakes that were missing Westland section of the Alpine Fault. from the Hokuri Creek record. We collected eight core samples to a Our new findings, based on the combined depth of seven metres from the wetland Hokuri Creek – John O’Groats records, immediately adjacent to the fault. By suggest there have been 27 previous examining these cores, we found evidence earthquakes, with a recurrence interval of past earthquakes in the form of peatof about 300 years (slightly shorter than silt couplets. Peat layers represent the However, the last thousand years of the previous estimate of 330 years). We wetland under stable conditions and silt sediment missing Hokuri Creek, so Whatiswe knewfrom before long earthquake records: have greater confidence in this revised layers represent deposition triggered by We are here the long record relies on comparisons with record because we have added to the Time (years) earthquakes. Radiocarbon dating of these past findings from Haast, 100 km away. Hokuri Creek data with a site in closer 4000 layers provided 6000 5000 3000 2000ages for seven 1000 1000earthquakes BC/AD 2000 This possible weakness led us in search of proximity. The new data will be fed into occurring within the last 2000 years. more clues from the nearby John O’Groats updated seismic hazard estimates that take River site (image, left). account of the wide variability in earthquake Haast in Earthquakes recurrence this record. New Zealand’s longest earthquake record consists of 24 earthquakes occurring over 8,000 years on the South-Westland section of the Alpine Fault at Hokuri Creek. This record provides a reasonable number of earthquakes on which to base hazard calculations and reveals that this part of the fault ruptures relatively regularly, with an average time between earthquakes of about 330 years.

What we know now:

We are here

Time (years) 6000

5000

4000

3000

2000

1000

Hokuri Creek Earthquakes

BC/AD

1000

2000

John O’Groats Earthquakes

The long earthquake record in the panel above clearly illustrates that another earthquake is inevitable and provides a much stronger dataset from which to estimate the time of next occurrence. 7


Ursula Cochran (left) and Kate Clark (right) examine a geological core for evidence of past earthquakes. Photo: Margaret Low, GNS Science

Size and Extent of Quakes We know these past earthquakes were large (magnitude 8 or greater) because they ruptured the ground surface, and we have found examples of large movements across the fault from a single event; 7.5 metres horizontally and 1 metre vertically. However, it is important to note that these results are just from the South-Westland section of the Alpine Fault. While many of the earthquakes recorded at these two sites are likely to have ruptured the central and possibly North-Westland sections of the fault as well, we don’t know the length of rupture from this research. Long earthquake records from lakes along the length of the Alpine Fault are proving very useful for determining which sections of the fault ruptured in which earthquakes.

8

Improve Our Understanding of the Hazards The last major earthquake on the Alpine Fault occurred in 1717. With an average of about 300 years between earthquakes, our findings suggest that we are due for another Alpine Fault earthquake in the near future. Simple actions, such as storing food and water, and securing large and heavy items, go a long way to prepare for such events, but also consider your neighbours, businesses, and wider community in your action plan. To learn more about what you can do to prepare for an earthquake, please visit ‘Get Ready, Get Thru.’

http://getthru.govt.nz/

With an average of about 300 years between earthquakes, our findings suggest that we are due for another Alpine Fault earthquake in the near future.

ALPINE FAULT

Earthquake geologists collecting cores in the John O’Groats wetland. Photo: GNS Science.

Researchers from the ‘Economics of Resilient Infrastructure’ research programme are working with roading and other infrastructure providers to better understand possible outcomes stemming from an Alpine Fault rupture and aftershock sequence. The ‘Economics of Resilient Infrastructure’ is funded by the Ministry of Business, Innovation and Employment.

9


UNDERSTANDING MODERN LIQUEFACTION TO UNDERSTAND THE PAST BY MONICA GIONA BUCCI (LINCOLN UNIVERSITY), PILAR VILLAMOR (GNS SCIENCE), PETER ALMOND (LINCOLN UNIVERSITY) Contact: [email protected]

During the Canterbury earthquake sequence, liquefaction was widespread and particularly prevalent in alluvial and coastal dune settings. We are using field and laboratory techniques to better understand the liquefaction susceptibility of these environments. Our aim is to build up a body of knowledge that will help us locate sites for studies of ancient liquefaction (paleoliquefaction).

10

UNDERSTANDING LIQUEFACTION

11


Right: Paleoseismic trench with evidence of blisters and sand volcanoes following the 14 February 2016 earthquake. Note the ground uplift on the left and the complementary depression on the right. Previous: Peter Almond examining features in the trench.

Evidence of paleoliquefaction can inform us about the history and distribution of liquefaction and associated earthquakes over a long time-frame, and is useful when attempting to study hidden faults that have failed to rupture at the surface or are covered by thick sediment. This is well understood from studies overseas, but gained traction in New Zealand following the 2010/11 Canterbury earthquakes.

Building a Liquefaction Profile Our research set out to describe the sediments and sedimentary layering in alluvial and coastal dune settings. Their susceptibility to liquefaction was not surprising, as these are geologically young sites (less than 4000 years old), made up of sandy/silty-sand sediments that are water-saturated and have not undergone the aging process that increases their cohesion. Our field studies involve paleoseismic trenching. A trench is the best exploratory method as it provides a visual crosssection of the sedimentary features in context, boundaries are clearly exposed, and ‘weathering’ features distinguish new liquefaction from old. 12

Importantly, the trench reveals distinct soil layers. As vegetation develops on the surface, chemical processes between the organic matter and the soil create three distinct layers. As shown, the top layer (A; topsoil) typically contains plant and organic material; the middle layer (B) is ‘weathered’ and characterised by reduction-oxidation processes, and below that is a nonweathered mineral layer (C). Understanding these patterns is critical to interpreting the data from the trench. Our first study site was on the floodplain of the Halswell River in Greenpark, near Christchurch. We dug trenches, obtained

soil samples for microscopic analysis, and investigated the shallow subsurface. We uncovered old buried river channels that had been in-filled with sediment, sandy river bars on the inside of river bends, buried soils, forests and swamps. We also identified sand volcanoes left behind by the 2010/11 Canterbury earthquakes that allowed us to trace their source and determine precisely which layers liquefied. Our observations suggest that while some alluvial settings are highly susceptible to liquefaction − we could identify liquefaction events dating back 1000 years − the distinct soil layers mostly remain intact.

A

A B C An example of how liquefaction may disrupt the soil profile.

B C

UNDERSTANDING LIQUEFACTION

Same photo with overlay: pink depicts the 2016 topsoil; orange identifies a 2010/11 sand volcano; brown identifies the pre-2010/11 topsoil; blue identifies the 14 February 2016 liquefaction features; light blue at the surface identifies sand volcanoes from the 14 February 2016 earthquake.

The 2016 Valentine’s Day Earthquake

Soil Science Held the Clues!

Geotechnical Assessments Key to Land-Use Planning

Our investigations next moved to the coastal dune environment with data collected from paleoseismic trenches in Wainoni and Queen Elizabeth II (QEII) Parks, and more recently at Pines Beach. In the weeks prior to 14 February 2016, open trenches were being analysed at Pines Beach. After the earthquake, we were able to quickly return to observe the immediate effects of liquefaction at the surface and within the trenches. Unlike the alluvial environment, the coastal dune data were more challenging to interpret.

In the coastal dune setting we observed that the soil layers could be highly disrupted by liquefaction, as shown in the graphic. We carefully mapped and analysed the layers to piece together what was happening. We observed that Pines Beach is prone to forming complex structures that consist of collapse features, blisters and sand volcanoes. Sub-horizontal injection of sand into the organic layer is common, which not only caused blisters on the ground but also split and deformed the uppermost soil layers.

Our data show that both alluvial and coastal dune environments are vulnerable to liquefaction, with coastal dunes being more susceptible. We now know more about how liquefaction manifests in these environments and what to look out for. Our findings reinforce the need for detailed geomorphological and sedimentological assessments to inform national land-use planning and will contribute to international understanding of different sediment settings prone to liquefaction.

Evidence of paleoliquefaction can inform us about the history and distribution of liquefaction and associated earthquakes over a long time-frame 13


STUDENT PROFILE: MONICA GIONA BUCCI Monica’s thesis includes pioneering methods to analyse the spatial patterns of surface liquefaction in coastal and alluvial environments. One aspect focuses on the detailed architecture of sediments susceptible to liquefaction at the macroand micro-scale, the latter which allows her to distinguish liquefaction injections of different ages.

A Thin section sample from alluvial setting; B Sample detail showing the important difference between liquefaction and the soil matrix fabric.

14

NHRP RESEARCH

INTRODUCING THE PLATFORM’S NEW PROJECTS: PARTNER-LED Quantifying Exposure To Specific And Multiple Volcanic Hazards

Building Quake & People: A Serious Game Platform for Informing Life-Saving Strategies

Jonathan Procter (Massey University), in collaboration with University of Auckland, University of Canterbury, Victoria University of Wellington and GNS Science.

Vicente Gonzalez (University of Auckland), in collaboration with Opus Research and University of Canterbury.

»» Volcanic eruptions are rarely single event hazards. Instead, a majority of eruptions worldwide manifest as complex evolving sequences with unforeseen impacts. This 4 year project will develop the first multi-stage, multi-hazard eruption impact model for New Zealand’s volcanoes. The research will develop a quantitative data set for multi-stage, multi-hazard volcanic eruptions; develop a novel stochastic model for analysis and prediction; and translate the estimated hazard impacts into forecasts of damages & loss of capacity.

Tools & Knowledge to Improve New Zealand’s Long Term Resilience to Wind Storms Peter Cenek (Opus Research), in collaboration with University of Auckland, GNS Science and NIWA. »» The typical wind storm can generate $10-40 million NZD worth of insured damages. Their accumulated impact to both built environment and commercial interests rivals that of earthquakes because of their frequency and pervasiveness. With climate change, the frequency of severe wind storms is predicted to increase. This 4 year project will inform and improve (i) mitigation measures such as retrofit, land-use planning and building code enforcement; (ii) current procedures for determining design wind speeds; and (iiI) our understanding of how climate change will affect the nation’s wind vulnerability.

»» There are a number of shortcomings in evacuation planning and research. This project will develop a computer-based modelling framework able to assess occupant behaviour in buildings in the event of an earthquake. A prototype game will be developed and tested in two case study building types heavily used by the public – councils and hospitals. The findings will contribute to improved strategies for building evacuation leading to improved citizen safety.

Climate Change Impacts on Weather-Related Hazards Giovanni Coco (University of Auckland), in collaboration with University of Waikato, NIWA, Scripps Institute of Oceanography (USA), Universidad de Cantabria (Spain), University of Florida (USA), University of New South Wales (Australia). »» This 4 year project will develop a finer-scale resolution (